Google has announced the release of Lyria 3 Pro, an advanced music generation model that allows users to create up to three-minute-long tracks. This upgrade from the initial Lyria 3 model not only extends the length but also enhances creative control, enabling users to specify detailed components such as intros and bridges within a musical piece.

Lyria 3 Pro integrates seamlessly with Google’s Gemini app, making it accessible to paid subscribers. It will also be available in Google Vids for video editing enthusiasts and via ProducerAI, an acquisition made by Google this month. The company is further expanding its AI music offerings through Vertex AI, the Gemini API, and AI Studio.

During training, Google utilised data from partners and YouTube, ensuring that the model does not mimic artists directly but can draw broad inspiration from them if requested. All tracks generated by these models will be marked with a SynthID, indicating their artificial origin.

In an increasingly AI-driven music landscape, other major players like Spotify have introduced tools to help artists prevent misattribution of their work. Meanwhile, Deezer has launched features aimed at identifying AI-generated music more broadly. As these technologies evolve, the challenge will be balancing creativity with ethical considerations and copyright issues.